AI for security - friend or foe?
summary
Panelists claim AI already cuts alert backlog to zero at one company and drops mean time to triage to minutes, describes an open‑sourced “easy agents” cloud webhook that spins up agents to investigate alerts and produce reproducible Jupyter notebooks, and describes a Codex‑based code scanner plus an agentic browser “Comet.” They demonstrate agent-driven triage and patch PR creation, warn about data leakage and agent identity and urge sandboxing, egress controls and audit trails as immediate priorities.
AI-written summary · about
chapters (8)
- 0:00Introduction and panel setup
- 3:07AI for security overview
- 9:52AI as defender and force multiplier
- 23:11Build versus buy in security tooling
- 35:21AI as adversary weapon
- 37:47AI adoption mistakes in security
- 39:52Lightning round predictions
- 41:18Closing remarks and Q&A
related talks
39mAI as a Catalyst, Not a Replacement
28mAI for Everything After Code
22mFireside Chat - AI Security Vulnerabilities
19mAI Agents for Incidents: The Good, the Bad, and the Ugly
16mAI Agents for Incident Investigation: The Good, The Bad, and The Ugly
30mAI Agents for Incident Management
44mPannel Discussion | Agentic Automation: Are AI Agents The Future of Digital Work
24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…